10 files changed, 121 insertions, 135 deletions

diff --git a/bench/build.zig b/bench/build.zig
index 66864c2..b7fb6af 100644
--- a/bench/build.zig
+++ b/bench/build.zig
@@ -73,7 +73,7 @@ pub fn build(b: *std.Build) !void {
             .name = "zg_caseless",
             .src = "src/zg_caseless.zig",
             .imports = &.{
-                .{ .name = "CaseFold", .module = zg.module("CaseFold") },
+                .{ .name = "CaseFolding", .module = zg.module("CaseFolding") },
                 .{ .name = "Normalize", .module = zg.module("Normalize") },
             },
         },
@@ -88,7 +88,7 @@ pub fn build(b: *std.Build) !void {
             .name = "zg_grapheme",
             .src = "src/zg_grapheme.zig",
             .imports = &.{
-                .{ .name = "grapheme", .module = zg.module("grapheme") },
+                .{ .name = "Graphemes", .module = zg.module("Graphemes") },
             },
         },
         .{
@@ -102,7 +102,7 @@ pub fn build(b: *std.Build) !void {
             .name = "zg_case",
             .src = "src/zg_case.zig",
             .imports = &.{
-                .{ .name = "CaseData", .module = zg.module("CaseData") },
+                .{ .name = "LetterCasing", .module = zg.module("LetterCasing") },
             },
         },
     };
@@ -129,15 +129,15 @@ pub fn build(b: *std.Build) !void {
         .target = target,
         .optimize = optimize,
     });
-    unit_tests.root_module.addImport("GenCatData", zg.module("GenCatData"));
-    unit_tests.root_module.addImport("PropsData", zg.module("PropsData"));
-    unit_tests.root_module.addImport("CaseData", zg.module("CaseData"));
+    unit_tests.root_module.addImport("GeneralCategories", zg.module("GeneralCategories"));
+    unit_tests.root_module.addImport("Properties", zg.module("Properties"));
+    unit_tests.root_module.addImport("LetterCasing", zg.module("LetterCasing"));
     unit_tests.root_module.addImport("Normalize", zg.module("Normalize"));
-    unit_tests.root_module.addImport("CaseFold", zg.module("CaseFold"));
+    unit_tests.root_module.addImport("CaseFolding", zg.module("CaseFolding"));
     unit_tests.root_module.addImport("DisplayWidth", zg.module("DisplayWidth"));
     unit_tests.root_module.addImport("code_point", zg.module("code_point"));
-    unit_tests.root_module.addImport("grapheme", zg.module("grapheme"));
-    unit_tests.root_module.addImport("ScriptsData", zg.module("ScriptsData"));
+    unit_tests.root_module.addImport("Graphemes", zg.module("Graphemes"));
+    unit_tests.root_module.addImport("Scripts", zg.module("Scripts"));
 
     const run_unit_tests = b.addRunArtifact(unit_tests);
 
diff --git a/bench/build.zig.zon b/bench/build.zig.zon
index 2c56013..aec7f92 100644
--- a/bench/build.zig.zon
+++ b/bench/build.zig.zon
@@ -1,12 +1,13 @@
 .{
-    .name = "zg",
+    .name = .zg,
     .version = "0.0.0",
-    .minimum_zig_version = "0.12.0",
+    .minimum_zig_version = "0.14.0",
+    .fingerprint = 0x47df77786eba1512,
 
     .dependencies = .{
         .ziglyph = .{
-            .url = "https://codeberg.org/dude_the_builder/ziglyph/archive/main.tar.gz",
-            .hash = "12207831bce7d4abce57b5a98e8f3635811cfefd160bca022eb91fe905d36a02cf25",
+            .url = "https://codeberg.org/dgv/ziglyph/archive/main.tar.gz",
+            .hash = "ziglyph-0.11.3-o3ARBKTuHwARGJBNZfNXfZsxxS1wCEbI745rgjYnyk4c",
         },
         .zg = .{ .path = ".." },
     },
diff --git a/bench/run_benches.sh b/bench/run_benches.sh
index 0210622..e2a26a8 100755
--- a/bench/run_benches.sh
+++ b/bench/run_benches.sh
@@ -1,6 +1,9 @@
 #!/bin/sh
 
-zig build -Doptimize=ReleaseSafe
+if ! zig build -Doptimize=ReleaseSafe; then
+  exit 1
+fi
+
 cd zig-out/bin/
 
 echo "\nBinary sizes ======="
diff --git a/bench/src/tests.zig b/bench/src/tests.zig
index a8a2a98..cf62709 100644
--- a/bench/src/tests.zig
+++ b/bench/src/tests.zig
@@ -6,140 +6,133 @@ const expectEqualStrings = testing.expectEqualStrings;
 
 const allocator = testing.allocator;
 
-const GenCatData = @import("GenCatData");
-
-test "General Category" {
-    const gcd = try GenCatData.init(allocator);
-    defer gcd.deinit();
-
-    try expect(gcd.gc('A') == .Lu); // Lu: uppercase letter
-    try expect(gcd.gc('3') == .Nd); // Nd: Decimal number
-    try expect(gcd.isControl(0));
-    try expect(gcd.isLetter('z'));
-    try expect(gcd.isMark('\u{301}'));
-    try expect(gcd.isNumber('3'));
-    try expect(gcd.isPunctuation('['));
-    try expect(gcd.isSeparator(' '));
-    try expect(gcd.isSymbol('©'));
+const GeneralCategories = @import("GeneralCategories");
+
+test GeneralCategories {
+    const gen_cat = try GeneralCategories.init(allocator);
+    defer gen_cat.deinit(allocator);
+
+    try expect(gen_cat.gc('A') == .Lu); // Lu: uppercase letter
+    try expect(gen_cat.gc('3') == .Nd); // Nd: Decimal number
+    try expect(gen_cat.isControl(0));
+    try expect(gen_cat.isLetter('z'));
+    try expect(gen_cat.isMark('\u{301}'));
+    try expect(gen_cat.isNumber('3'));
+    try expect(gen_cat.isPunctuation('['));
+    try expect(gen_cat.isSeparator(' '));
+    try expect(gen_cat.isSymbol('©'));
 }
 
-const PropsData = @import("PropsData");
+const Properties = @import("Properties");
 
-test "Properties" {
-    const pd = try PropsData.init(allocator);
-    defer pd.deinit();
+test Properties {
+    const props = try Properties.init(allocator);
+    defer props.deinit(allocator);
 
-    try expect(pd.isMath('+'));
-    try expect(pd.isAlphabetic('Z'));
-    try expect(pd.isWhitespace(' '));
-    try expect(pd.isHexDigit('f'));
-    try expect(!pd.isHexDigit('z'));
+    try expect(props.isMath('+'));
+    try expect(props.isAlphabetic('Z'));
+    try expect(props.isWhitespace(' '));
+    try expect(props.isHexDigit('f'));
+    try expect(!props.isHexDigit('z'));
 
-    try expect(pd.isDiacritic('\u{301}'));
-    try expect(pd.isIdStart('Z')); // Identifier start character
-    try expect(!pd.isIdStart('1'));
-    try expect(pd.isIdContinue('1'));
-    try expect(pd.isXidStart('\u{b33}')); // Extended identifier start character
-    try expect(pd.isXidContinue('\u{e33}'));
-    try expect(!pd.isXidStart('1'));
+    try expect(props.isDiacritic('\u{301}'));
+    try expect(props.isIdStart('Z')); // Identifier start character
+    try expect(!props.isIdStart('1'));
+    try expect(props.isIdContinue('1'));
+    try expect(props.isXidStart('\u{b33}')); // Extended identifier start character
+    try expect(props.isXidContinue('\u{e33}'));
+    try expect(!props.isXidStart('1'));
 
     // Note surprising Unicode numeric types!
-    try expect(pd.isNumeric('\u{277f}'));
-    try expect(!pd.isNumeric('3'));
-    try expect(pd.isDigit('\u{2070}'));
-    try expect(!pd.isDigit('3'));
-    try expect(pd.isDecimal('3'));
+    try expect(props.isNumeric('\u{277f}'));
+    try expect(!props.isNumeric('3'));
+    try expect(props.isDigit('\u{2070}'));
+    try expect(!props.isDigit('3'));
+    try expect(props.isDecimal('3'));
 }
 
-const CaseData = @import("CaseData");
+const LetterCasing = @import("LetterCasing");
 
-test "Case" {
-    const cd = try CaseData.init(allocator);
-    defer cd.deinit();
+test LetterCasing {
+    const case = try LetterCasing.init(allocator);
+    defer case.deinit(allocator);
 
-    try expect(cd.isUpper('A'));
-    try expect('A' == cd.toUpper('a'));
-    try expect(cd.isLower('a'));
-    try expect('a' == cd.toLower('A'));
+    try expect(case.isUpper('A'));
+    try expect('A' == case.toUpper('a'));
+    try expect(case.isLower('a'));
+    try expect('a' == case.toLower('A'));
 
-    try expect(cd.isCased('É'));
-    try expect(!cd.isCased('3'));
+    try expect(case.isCased('É'));
+    try expect(!case.isCased('3'));
 
-    try expect(cd.isUpperStr("HELLO 123!"));
-    const ucased = try cd.toUpperStr(allocator, "hello 123");
+    try expect(case.isUpperStr("HELLO 123!"));
+    const ucased = try case.toUpperStr(allocator, "hello 123");
     defer allocator.free(ucased);
     try expectEqualStrings("HELLO 123", ucased);
 
-    try expect(cd.isLowerStr("hello 123!"));
-    const lcased = try cd.toLowerStr(allocator, "HELLO 123");
+    try expect(case.isLowerStr("hello 123!"));
+    const lcased = try case.toLowerStr(allocator, "HELLO 123");
     defer allocator.free(lcased);
     try expectEqualStrings("hello 123", lcased);
 }
 
 const Normalize = @import("Normalize");
 
-test "Normalization" {
-    var norm_data = try Normalize.NormData.init(allocator);
-    defer norm_data.deinit();
-    const n = Normalize{ .norm_data = &norm_data };
+test Normalize {
+    const normalize = try Normalize.init(allocator);
+    defer normalize.deinit(allocator);
 
     // NFD: Canonical decomposition
-    const nfd_result = try n.nfd(allocator, "Héllo World! \u{3d3}");
-    defer nfd_result.deinit();
+    const nfd_result = try normalize.nfd(allocator, "Héllo World! \u{3d3}");
+    defer nfd_result.deinit(allocator);
     try expectEqualStrings("He\u{301}llo World! \u{3d2}\u{301}", nfd_result.slice);
 
     // NFKD: Compatibility decomposition
-    const nfkd_result = try n.nfkd(allocator, "Héllo World! \u{3d3}");
-    defer nfkd_result.deinit();
+    const nfkd_result = try normalize.nfkd(allocator, "Héllo World! \u{3d3}");
+    defer nfkd_result.deinit(allocator);
     try expectEqualStrings("He\u{301}llo World! \u{3a5}\u{301}", nfkd_result.slice);
 
     // NFC: Canonical composition
-    const nfc_result = try n.nfc(allocator, "Complex char: \u{3D2}\u{301}");
-    defer nfc_result.deinit();
+    const nfc_result = try normalize.nfc(allocator, "Complex char: \u{3D2}\u{301}");
+    defer nfc_result.deinit(allocator);
     try expectEqualStrings("Complex char: \u{3D3}", nfc_result.slice);
 
     // NFKC: Compatibility composition
-    const nfkc_result = try n.nfkc(allocator, "Complex char: \u{03A5}\u{0301}");
-    defer nfkc_result.deinit();
+    const nfkc_result = try normalize.nfkc(allocator, "Complex char: \u{03A5}\u{0301}");
+    defer nfkc_result.deinit(allocator);
     try expectEqualStrings("Complex char: \u{038E}", nfkc_result.slice);
 
     // Test for equality of two strings after normalizing to NFC.
-    try expect(try n.eql(allocator, "foé", "foe\u{0301}"));
-    try expect(try n.eql(allocator, "foϓ", "fo\u{03D2}\u{0301}"));
+    try expect(try normalize.eql(allocator, "foé", "foe\u{0301}"));
+    try expect(try normalize.eql(allocator, "foϓ", "fo\u{03D2}\u{0301}"));
 }
 
-const CaseFold = @import("CaseFold");
+const CaseFolding = @import("CaseFolding");
 
-test "Caseless matching" {
-    var norm_data = try Normalize.NormData.init(allocator);
-    defer norm_data.deinit();
-    const n = Normalize{ .norm_data = &norm_data };
-
-    const cfd = try CaseFold.FoldData.init(allocator);
-    defer cfd.deinit();
-    const cf = CaseFold{ .fold_data = &cfd };
+test CaseFolding {
+    const case_fold = try CaseFolding.init(allocator);
+    defer case_fold.deinit(allocator);
 
     // compatCaselessMatch provides the deepest level of caseless
     // matching because it decomposes and composes fully to NFKC.
     const a = "Héllo World! \u{3d3}";
     const b = "He\u{301}llo World! \u{3a5}\u{301}";
-    try expect(try cf.compatCaselessMatch(allocator, &n, a, b));
+    try expect(try case_fold.compatCaselessMatch(allocator, a, b));
 
     const c = "He\u{301}llo World! \u{3d2}\u{301}";
-    try expect(try cf.compatCaselessMatch(allocator, &n, a, c));
+    try expect(try case_fold.compatCaselessMatch(allocator, a, c));
 
     // canonCaselessMatch isn't as comprehensive as compatCaselessMatch
     // because it only decomposes and composes to NFC. But it's faster.
-    try expect(!try cf.canonCaselessMatch(allocator, &n, a, b));
-    try expect(try cf.canonCaselessMatch(allocator, &n, a, c));
+    try expect(!try case_fold.canonCaselessMatch(allocator, a, b));
+    try expect(try case_fold.canonCaselessMatch(allocator, a, c));
 }
 
 const DisplayWidth = @import("DisplayWidth");
 
-test "Display width" {
-    const dwd = try DisplayWidth.DisplayWidthData.init(allocator);
-    defer dwd.deinit();
-    const dw = DisplayWidth{ .data = &dwd };
+test DisplayWidth {
+    const dw = try DisplayWidth.init(allocator);
+    defer dw.deinit(allocator);
 
     // String display width
     try expectEqual(@as(usize, 5), dw.strWidth("Hello\r\n"));
@@ -197,13 +190,13 @@ test "Code point iterator" {
     }
 }
 
-const grapheme = @import("grapheme");
+const Graphemes = @import("Graphemes");
 
 test "Grapheme cluster iterator" {
-    const gd = try grapheme.GraphemeData.init(allocator);
-    defer gd.deinit();
+    const graphemes = try Graphemes.init(allocator);
+    defer graphemes.deinit(allocator);
     const str = "He\u{301}"; // Hé
-    var iter = grapheme.Iterator.init(str, &gd);
+    var iter = graphemes.iterator(str);
     var i: usize = 0;
 
     while (iter.next()) |gc| : (i += 1) {
@@ -217,13 +210,13 @@ test "Grapheme cluster iterator" {
     }
 }
 
-const ScriptsData = @import("ScriptsData");
+const Scripts = @import("Scripts");
 
-test "Scripts" {
-    const sd = try ScriptsData.init(allocator);
-    defer sd.deinit();
+test Scripts {
+    const scripts = try Scripts.init(allocator);
+    defer scripts.deinit(allocator);
 
-    try expect(sd.script('A') == .Latin);
-    try expect(sd.script('Ω') == .Greek);
-    try expect(sd.script('צ') == .Hebrew);
+    try expect(scripts.script('A') == .Latin);
+    try expect(scripts.script('Ω') == .Greek);
+    try expect(scripts.script('צ') == .Hebrew);
 }
diff --git a/bench/src/zg_case.zig b/bench/src/zg_case.zig
index 7308e82..c444343 100644
--- a/bench/src/zg_case.zig
+++ b/bench/src/zg_case.zig
@@ -1,6 +1,6 @@
 const std = @import("std");
 
-const CaseData = @import("CaseData");
+const LetterCasing = @import("LetterCasing");
 
 pub fn main() !void {
     var arena = std.heap.ArenaAllocator.init(std.heap.page_allocator);
@@ -18,15 +18,15 @@ pub fn main() !void {
     );
     defer allocator.free(input);
 
-    const case_data = try CaseData.init(allocator);
+    const case = try LetterCasing.init(allocator);
 
     var iter = std.mem.splitScalar(u8, input, '\n');
     var result: usize = 0;
     var timer = try std.time.Timer.start();
 
     while (iter.next()) |line| {
-        const upper = try case_data.toUpperStr(allocator, line);
-        const lower = try case_data.toLowerStr(allocator, line);
+        const upper = try case.toUpperStr(allocator, line);
+        const lower = try case.toLowerStr(allocator, line);
         result += upper.len + lower.len;
     }
     std.debug.print("zg toUpperStr/toLowerStr: result: {}, took: {}\n", .{ result, std.fmt.fmtDuration(timer.lap()) });
@@ -36,8 +36,8 @@ pub fn main() !void {
     timer.reset();
 
     while (iter.next()) |line| {
-        if (case_data.isUpperStr(line)) result += 1;
-        if (case_data.isLowerStr(line)) result += 2;
+        if (case.isUpperStr(line)) result += 1;
+        if (case.isLowerStr(line)) result += 2;
     }
     std.debug.print("zg isUpperStr/isLowerStr: result: {}, took: {}\n", .{ result, std.fmt.fmtDuration(timer.lap()) });
 }
diff --git a/bench/src/zg_caseless.zig b/bench/src/zg_caseless.zig
index fa350e9..5d8d591 100644
--- a/bench/src/zg_caseless.zig
+++ b/bench/src/zg_caseless.zig
@@ -1,7 +1,6 @@
 const std = @import("std");
 
-const CaseFold = @import("CaseFold");
-const Normalize = @import("Normalize");
+const CaseFolding = @import("CaseFolding");
 
 pub fn main() !void {
     var arena = std.heap.ArenaAllocator.init(std.heap.page_allocator);
@@ -19,11 +18,7 @@ pub fn main() !void {
     );
     defer allocator.free(input);
 
-    const fold_data = try CaseFold.FoldData.init(allocator);
-    var case_fold = CaseFold{ .fold_data = &fold_data };
-    var norm_data: Normalize.NormData = undefined;
-    try Normalize.NormData.init(&norm_data, allocator);
-    var normalize = Normalize{ .norm_data = &norm_data };
+    const case_fold = try CaseFolding.init(allocator);
 
     var iter = std.mem.splitScalar(u8, input, '\n');
     var result: usize = 0;
@@ -34,14 +29,13 @@ pub fn main() !void {
     while (iter.next()) |line| {
         if (try case_fold.compatCaselessMatch(
             allocator,
-            &normalize,
             prev_line,
             line,
         )) result += 1;
         @memcpy(buf[0..line.len], line);
         prev_line = buf[0..line.len];
     }
-    std.debug.print("zg CaseFold.compatCaselessMatch: result: {}, took: {}\n", .{ result, std.fmt.fmtDuration(timer.lap()) });
+    std.debug.print("zg CaseFolding.compatCaselessMatch: result: {}, took: {}\n", .{ result, std.fmt.fmtDuration(timer.lap()) });
 
     result = 0;
     iter.reset();
@@ -50,12 +44,11 @@ pub fn main() !void {
     while (iter.next()) |line| {
         if (try case_fold.canonCaselessMatch(
             allocator,
-            &normalize,
             prev_line,
             line,
         )) result += 1;
         @memcpy(buf[0..line.len], line);
         prev_line = buf[0..line.len];
     }
-    std.debug.print("zg CaseFold.canonCaselessMatch: result: {}, took: {}\n", .{ result, std.fmt.fmtDuration(timer.lap()) });
+    std.debug.print("zg CaseFolding.canonCaselessMatch: result: {}, took: {}\n", .{ result, std.fmt.fmtDuration(timer.lap()) });
 }
diff --git a/bench/src/zg_codepoint.zig b/bench/src/zg_codepoint.zig
index 7a8ccbe..2fe31f3 100644
--- a/bench/src/zg_codepoint.zig
+++ b/bench/src/zg_codepoint.zig
@@ -1,6 +1,6 @@
 const std = @import("std");
 
-const CodePointIterator = @import("code_point").Iterator;
+const code_point = @import("code_point");
 
 pub fn main() !void {
     var arena = std.heap.ArenaAllocator.init(std.heap.page_allocator);
@@ -18,7 +18,7 @@ pub fn main() !void {
     );
     defer allocator.free(input);
 
-    var iter = CodePointIterator{ .bytes = input };
+    var iter = code_point.Iterator{ .bytes = input };
     var result: usize = 0;
     var timer = try std.time.Timer.start();
 
diff --git a/bench/src/zg_grapheme.zig b/bench/src/zg_grapheme.zig
index 74fb8b5..ee40de5 100644
--- a/bench/src/zg_grapheme.zig
+++ b/bench/src/zg_grapheme.zig
@@ -1,7 +1,6 @@
 const std = @import("std");
 
-const GraphemeData = @import("grapheme").GraphemeData;
-const GraphemeIterator = @import("grapheme").Iterator;
+const Graphemes = @import("Graphemes");
 
 pub fn main() !void {
     var arena = std.heap.ArenaAllocator.init(std.heap.page_allocator);
@@ -19,11 +18,11 @@ pub fn main() !void {
     );
     defer allocator.free(input);
 
-    const grapheme_data = try GraphemeData.init(allocator);
-    var iter = GraphemeIterator.init(input, &grapheme_data);
+    const graphemes = try Graphemes.init(allocator);
+    var iter = graphemes.iterator(input);
     var result: usize = 0;
     var timer = try std.time.Timer.start();
 
     while (iter.next()) |_| result += 1;
-    std.debug.print("zg GraphemeIterator: result: {}, took: {}\n", .{ result, std.fmt.fmtDuration(timer.lap()) });
+    std.debug.print("zg Graphemes.Iterator: result: {}, took: {}\n", .{ result, std.fmt.fmtDuration(timer.lap()) });
 }
diff --git a/bench/src/zg_normalize.zig b/bench/src/zg_normalize.zig
index 1e2cfab..268c060 100644
--- a/bench/src/zg_normalize.zig
+++ b/bench/src/zg_normalize.zig
@@ -18,9 +18,7 @@ pub fn main() !void {
     );
     defer allocator.free(input);
 
-    var norm_data: Normalize.NormData = undefined;
-    try Normalize.NormData.init(&norm_data, allocator);
-    const normalize = Normalize{ .norm_data = &norm_data };
+    const normalize = try Normalize.init(allocator);
 
     var iter = std.mem.splitScalar(u8, input, '\n');
     var result: usize = 0;
diff --git a/bench/src/zg_width.zig b/bench/src/zg_width.zig
index 041d2c5..b2db3fa 100644
--- a/bench/src/zg_width.zig
+++ b/bench/src/zg_width.zig
@@ -18,8 +18,7 @@ pub fn main() !void {
     );
     defer allocator.free(input);
 
-    const width_data = try DisplayWidth.DisplayWidthData.init(allocator);
-    const display_width = DisplayWidth{ .data = &width_data };
+    const display_width = try DisplayWidth.init(allocator);
 
     var iter = std.mem.splitScalar(u8, input, '\n');
     var result: usize = 0;